RELATION OF DIAGENETIC ALTERATION AND SOIL- FORMING PROCESSES TO THE URANIUM DEPOSITS

OF THE SOUTHEAST TEXAS COASTAL PLAIN1

ALICE D. WEEKS AND D. HOYE EAR~LE U.S. Geological Survey, Washington, D.C., and Austin, Texas

ABSTRACT

The Upper Eocene Jackson Group is the chief host rock of the uranium deposits in the Kames area of the southeast Texas Coastal Plain. It is highly tuffaceous and the uranium deposits are within approximately 100 f t of the unconformity with the overlying Cata- houla Tuff (Miocene ?). Glass shards, fragments of sanidine and plagioclase, grains of fine- grained volcanic rocks, and biotite and other minerals in these sediments were highly reactive chemically, causing complex diagenetic alteration and the development of alkaline carbonate pore water. Extensive caliche development and silica induration associated with a recent hotter, drier climate favored the concentration of uranium. The origin of these shallow uranium deposits is believed to have been controlled by the complex diagenesis of the highly reactive volcanic detritus, by development of a "built- in" solvent for uranium (the alkaline carbonate pore water), and by climatic, structural, and permeability conditions that allowed concentration and deposition of uranium rather than dilution and dispersal. The tuffaceous rocks are considered to have been the source of the uranium and associated molybdenum, phosphorus, and arsenic.

INTRODUCTION

The purpose of this paper is to outline the important relations between diagenetic and soil-forming processes in tuffaceous rocks and the origin of the uranium deposits in the southeast Texas Coastal Plain (Fig. 1). This discussion supplements the description of the uranium deposits given in the guidebook of the Coastal Plain field trip conducted as part of the program of the Tenth National Clay Conference (Eargleand Weeks, 196laf. The climate of the area, the character of the ground water, and the geologic structure are significant factors in the geochemical environment and origin of the deposits. I t seems to us that the devdopment of the uranium deposits in the Tertiary sedimentary rocks of the Texas Coastal Plain began with the diagenesis of the sediments and continued for millions of years. Modification of the deposits by soil-forming processes is still going on.

Publication authorized by the Director, U.S. Geological Survey. 23

24 TE~T~ ~NATIONAL CON~'ERENc/~ ON (~.L,Au AND CLAY MINEI~ALS

o ~ ~- c ~ ~

,. ,-~ Z 4--

W

~,__~ d~1oJs au-~oq ! o l d 3 \

auavo[i d auaoo !v 4 4-05!10 ~ t J a 3 o ]

Lq-

o ~ (/3

~s

r

O O

l.

DIAGt~:NETIC ALTERATION AND ~OILoFORMI]NG PROCESSES 25

M E A N I N G OF D I A G E N E S I S

Possibly because we define diagenesis broadly, we have trouble finding a break between the " e n d " of diagenesis and the "beginning" of uranium mineralization. Dapples (1959) has used the term diagenesis " in an informal sense to designate the modifications which a sediment experiences during its deposition and lithification". He distinguished three stages in the diagenesis of quartzose sandstones : (1) deposition characterized by rounding and pitting of quartz grains and instability of some minerals with minor but persistent solution of silica, (2) early burial characterized by precipitation of quartz as overgrowths, and (3) late burial characterized by addition of carbonate as cement and as a replacement of quartz. Some geologists would limit dia- genesis to alterations which occur prior to cementation and others would include any transformation before metamorphism. Pett i john (1957, p. 648) defined diagenesis as " the reactions which take place within a sediment between one mineral and another, or between one or more minerals and the interstitial or supernatant fluids ", and considered the processes involved to include cementation, authigenesis, diagenetic differentiation and segregation, diagenetic metasomatism, intrastratal solution, and compaction. The latter definition is more inclusive and more applicable to the diagenesis of the tuffaceous rocks in the Karnes area of Texas.

C H A R A C T E R OF U R A N I U M H O S T R O C K S

In general, the host rocks of the so-called sandstone-type uranium deposits (outside of Texas as well as in Texas) are quite different geochemically from average marine quartzosc sandstones, and their diagenesis is more complex. The uranium deposits are mainly in terrestrial sediments, representing alluvial plains, basin fill, coastal plains, and some lacustrine deposits. The host rocks are chiefly tuffaceous or arkosie sandstone with some conglomerate, shale, lignite, and limestone or sandstone closely associated with tuffaceous rocks. The rocks are characterized by silicic, calcitic, or dolomitic cements and by various types of clay and ehloritic alteration products (Weeks, Truesdell and Haffty, 1957). Some show the development of one or more zeolitic minerals. The sediments have local areas of reducing environment associated with pyrite or other sulfide minerals, coalified wood, humic material, H2S, or asphaltic(?) material.

In particular, the Upper Eocene Jackson Group of the Karncs area in the southeast Texas Coastal Plain has alternating marine and nonmarine sedi- ments in thin lithologic units tha t include highly feldspathic and lithie (volcanic) sandstone, tuffaceous sand, shardy ash, lignitic clay, and marine fossiliferous sandy clays. The Catahoula Tuff, of Miocene(?) age and uncon- formable on the Jackson, consists of thick ash, tuffaceous sand, and thin conglomerates of volcanic rock fragments including andesite, trachyandesite, trachyte, rhyolite, and pumice. The tuffaceous, feldspathie and lithic sedi- ments are unstable in the sedimentary environment and are much more

26 TENTH NATIONAL CONFERENCE ON CLAYS AND CLAY MINERALS

highly reactive chemically than normal marine sands and lagoonal clays, which consist chiefly of relatively insoluble quartz and clay minerals with traces of resistant heavy minerals. The volcanic shards are particularly unstable because of their glassy state and large surface area. The sand-size grains in the Jackson include many cleavage fragments of sanidine and zoned plagio- clase phenocrysts, and grains of aphanitic to trachytic fine-grained volcanic rocks, with traces of "books" of biotite and other marie minerals. In general, these are much more susceptible to alteration than is quartz.

D I A G E N E S I S I N T H E K A t ~ N E S A R E A

Several types of diagenetic alteration and a variety of authigenic minerals have been noted in tuffaceous rocks of the Jackson Group. The alteration of tufts to montmorillonitic clay is an irregular or nonuniform process, inas- much as some layers have altered to fairly well-developed bentonite, and other layers are slightly altered with fair to good shardy texture still pre- served. In the Karnes area all the parts of the Jackson Group, but not the Catahoula Tuft, show some zeolitic alteration with development of authigenic heulandite (probably the variety clinoptilolite) in microscopic interstitial crystals, or as a firm cementing material in the feldspathic and lithic sand- stones; also the zeolite is present in fine-grained altered tufts where it was detected only by an X-ray diffraction pattern. What seem to be authigenic feldspar overgrowths are found in several thin sections of feldspathic sand- stones. The silica that was released in the alteration of volcanic glass formed much opal and chalcedony both as cements in the sandstones and as opalized wood and tuft. The pore water became saturated with silica from the volcanic glass before any appreciable etching of quartz had taken place. Etching of quartz grains is important only locally, and redeposition of silica as quartz overgrowths is minor compared to the amount of opaline coating and ehalce- donic interstitial filling.

Diagenetie alteration also includes the migration of manganese and its redeposition as oxide in spots and seams. Locally in the Lyssy (San Antonio Mining Company) prospect trench in western Karnes County, seams were as much as ~ in. thick. Fine-grained pyrite and marcasite are commonly, but not necessarily, associated with plant fragments or other organic material. Siliceous or calcareous concretions occur in several zones in the upper part of the Jackson Group; in particular, one layer in the Stones Switch Sandstone Member of the Whitsett Formation in the Luekett mine in western Karnes County has dense calcareous concretions several feet in diameter. Authigenic barite was noted in many heavy-mineral fractions, and siderite is present in a few samples.

A sharp distinction cannot be drawn between the authigenic minerals formed by diagenetic alteration and the uranium ore and gangue minerals. Neither can a specific paragenetic sequence of ore and gangue minerals be established for this type of shallow uranium deposit because several of the minerals probably formed at more than one t ime (see discussion of origin).

])IAG]~I~ETIC ALTERATION AND SOIL-FORMING ~I%OCESSES 27

WEATHERING AND SOIL-FORMING PROCESSES IN KARNES AREA

The Karnes area is generally soil-covered except for natural exposures of bare rock along the frontal slopes of the steeper cnestas and along banks undercut by streams. The weathering and soil-ferruling processes, conditioned by the climate, have played an important role in the migration and redeposi- tion of uranium. The Karnes area is on the border between subhumid climate to the east and northeast and semiarid climate to the west and southwest. There is evidence of a change, since perhaps Middle Pleistocene time, from a hotter, drier climate with extensive caliche development and silica induration to a more humid climate with development of an organic-rich topsoil and with less ealiche in the subsoil. In humid areas most of the soluble products are flushed out and removed from the area by ground and sm'faee drainage. In drier climates soluble materials arc removed to a lesser degree, depending on the amount of accumulation during hot, dry weather compared to the amount lost during occasional heavy rains; also soil and ground water tend to become alkaline and high in carbonate.

Of interest in relation to the uranium deposits are the caliehe development and the silieification. In the Idarnes area, the thick ealiche layer of a former drier climate is largely removed; much has been redistributed as veinlets in the subsoil, but some fairly continuous patches remain on the outcrop of the Catahoula Tuff. This area is about the northeastern limit of a large area of caliche which extends southwest and south through I)uval County, Texas, into Mexico. It is also the northeastern limit (as far as known in 1961) of commercial uranium deposits in the Texas Coastal Plain. The carbonate ground water from which the caliehe developed also could have held the uranium in solution as uranyl carbonate complex ions. The warm, dry conditions that caused the caliehe accumulation through evaporation and loss of CO 2 also favored concentration rather than dispersal of uranium. Caliehe commonly forms by deposition of calcium carbonate derived from limestone or carbonate minerals; calcium carbonate can also develop in altered dacitic or andesitic tufts by the release of calcium from volcanic glass or from calcium-bearing minerals and pick-up of carbon dioxide from the air.

As the tuffaceous sediments are altered, considerable silica is released. It may be transported only a few feet or perhaps much farther before forming opalized wood or opalized ash, or coating dctrital grains in sandstone. Many silicified sandstone beds are found along the outcrop of the Jackson Group and of the Catahoula Tuff in Karnes County and a short distance southwest into Ataseosa County. Fine-grained sandstone commonly has thin rims of opal on the sand grains and medium-grained sandstone generally has alternat- ing opal and chalcedony bands. A few samples show an interstitial fill chiefly of chalcedony with very little opal. This silicification is a surface phenomenon that rarely extends more than l0 ft below the surface, as seen ill many of the uranium prospect trenches as well as in drill cores. It is not continuous at the surface, but gives the impression of being related to a slightly older erosion

28 TECTal. NATJ.ONAL CONFERENCE O~ CLAYS AI~D CLAY MII~ERALS

surface tha t is now being gently dissected and from which the caliche cap has been removed. At most of the uranium prospect trenches and opeu-pit mines the silicified rock has little or no uranium content. The ore is usually in friable sand under the silicified cap. In a few places friable sand of ore grade has been found close to the surface without the silicified cover, and at one prospect (Jacobs Ranch in southwestern Gonzales County) uranyl vanadate minerals were in a silicified coarse sandstone directly at the surface of the ground.

C H A R A C T E R OF T H E G R O U N D W A T E R

Recent publications on the ground-water geology of Karnes County (Anders, 1960), Wilson County (Anders, 1957), and Live Oak County (Anders and Baker, 1961) indicate certain characteristics of the ground water in formations of the Karnes area tha t are pertinent to our discussion of altered tufts and the origin of the uranium deposits. With some precautions, ground- water data may be interpreted to reflect the nature of the formation pore water--par t icular ly in this subhumid to semiarid area where many of the rocks are poor aquifers, recharge is very slow, and the wells are mostly of moderate depth. Only in the one good aquifer, the Carrizo Sand of Middle Eocene age, is there much recharge, and fresh water extends to a depth of several thousand feet.

In Wilson County just north of the uranium area, a small number of wells in the Jackson Group generally have rather highly mineralized water of poor quality. In Karnes County, a larger number of wells in the Jackson range from a small yield of 10 to 50 gal per min to moderate yield of 50 to 500 gal per rain. Quality ranges from " f re sh" (containing less than 1000 ppm total dissolved solids) to moderately saline (with total dissolved solids between 3000 and 10,000 ppm).

Of interest are the silica, sodium, bicarbonate, fluoride, boron, and pH. Only the more recent chemical analyses include the determination of silica, or about ]00 of the several hundred analyses in these reports (Anders, 1957, 1960; Anders and Baker, 1961). Figure 2 shows the silica content in ppm in water samples from the Lagarto Clay (Miocene ?) and Oakville Sandstone (Miocene ?) grouped together, the Catahoula Tuff (Miocene ?), the Jackson Group (Upper Eocene), the Yegua Formation (Middle Eocene), and the Carrizo Sand (Middle Eocene). I t is not possible to be certain that each water sample came exclusively from the formation indicated. To avoid confusion, samples known to be mixed Jackson and Catahoula water or Catahoula and Oakville water have been omit ted from the graph. Although the number of silica determinations for some of the formation waters is small, the Catahoula water samples tend to have much more silica than the others because of the large percentage of volcanic detritus in the Catahoula. Silica averages 80 ppm in the Catahoula samples and the largest number of samples are in the 90 to 100 ppm bracket, a content much higher than in most natural waters (Hem, 1959, pp. 56-57). The Jackson samples with an

DIAGENETIC ALTERATION AiND ~OIL-FOI~MllslG PROOJ~SSES 29

average of 49 ppm and the Oakvillo and Lagat~o samples with 46 ppm are notably less than the Catahoula but are still above the common 0 to 30 ppm range. The Jackson originally contained considerable volcanic material inter- bedded and partly mixed with nonvoleanic material. Its volcanic detritus is finer grained and more altered than that in the Catahoula, and more time

Cla F omd td O0"kvl l le

(41]) I ~ a n d s t o n e - - I I , L , ~ J i

0 50 tO0 ppm SiO z

ka

3 ~ 2 -

1 ~ Cafahoula

0 50 I00 pprn S;O~_

' - , ~ (91 Jo.ckson

0 50 I00 p p m $~o 2 i

a- w io

y e g u a 0 (9) ] i , i i , , m - ~ For rn o~-~on

o 50 I00 p p ~ S l 0 z

-i , C ' o r ~ ; z o ( ~ i i ~ 5 a n d

0 5 0 I00 p p m $!02.

]~'IQUI=~E 2.--Si l ica content in parts per million in ground-water samples from Karnes County and Live Oak County, Texas. (Analyses by the U.S. Geological Survey.)

has intervened since the Late Eocene for the formation of zeolite and authi- genie feldspar and the precipitation of silica in opalized wood and as interstitial fill. The interbedded nonvolcanie material--sand, shells, and wood--has afforded local sites of deposition for much of the silica of the Jackson. The 0akville Sandstone and Lagarto Clay contain less volcanic material than the Catahoula and are not much altered. The high silica in the water of the Karnes area is accompanied by high sodium, a common associa- tion (Hem, 1959, p. 57).

30 TEN~-[ NATIONAL CONFE~]~NC:E ON CLAYS AND CLAY MINERALS

The sodium content of the ground water is generally high and commonly Na+ is the dominant cation. Anders' (1960) report on Karnes Connty includes in the tables of water analyses several calculated parameters useful in judging the quality: percent sodium, sodium-absorption ratio (SAt~), and residual sodium carbonate. Nearly four-fifths of the samples have more than 50 percent sodium (with respect to total Na, Ca, Mg, and K), and the samples with more than 90 percent sodium are chiefly from the Catahoula or mixed Catahoula and Jackson waters. However, the average sodium content of the Jackson water samples is 675 ppm, somewhat above the average 533 ppm of Catahoula samples. This is par t ly because the Jackson includes more marine beds and its water contains more chloride than the Catahoula.

In many analyses potassium has been included with the sodium as equiva- lent sodium, but in those analyses for which potassium is listed separately, i t is commonly one-thirtieth to one-tenth as much as the sodium. Hem (1959, p. 91) states tha t the concentration of potassium seldom rises over 15 ppm in ordinary surface and ground water and is usually 10 ppm or less. In the Karnes County report only 57 potassium determinations are listed; the average potassium content is in the 25 to 30 ppm range for the Catahoula, Jackson, and Yegua samples, about 20 ppm for the Oakville, and less than 10 ppm for the Carrizo Sand and Lagarto Clay. The highest individual sample is 70 ppm from the Catahoula Tuff.

The high content of alkalies in the ground waters suggests tha t alkalies are being released by alteration of the volcanic glass grid the water has no~ moved far enough for the usual selective removal of potassium. The Cata- houla sediments probably have never been buried by more than a few hundred feet of overburden, which is too small a load to have squeezed out much pore water. In contrast, a few samples from a depth of several thousand feet in the Carrizo Sand have a potassium average of 7 ppm and sodium of 689 ppm. These waters have a high bicarbonate average of 1630 ppm, a small amount of carbonate (determined on only two samples) and p H ranging from 8.0 to 8.7. The range of p H for the Catahoula samples is 7.3 to 8.5 for 22 samples with an average of 7.73. The Jackson samples range in p H from 6.6 (only one sample below 7) to 8.3 with an average of 7.74. Other than the high bicarbo- nate samples of the Carrizo, only a few Jackson and Catahoula samples have a concentration of bicarbonate above 500 ppm (highest 831). Most of the Catahoula samples are in the 300 to 400 ppm range and average higher than the Jackson samples which show a wide spread and average 291 ppm. Silica tends to increase with increasing HCOs and alkalies up to a maximum controlled by the solubility and availability of silica. The Catahoula Tuff has the largest amount of available amorphous silica in the glass shards and the Carrizo Sand contains chiefly quartz with much lower solubility than glass, although the bicarbonate content of the Carrizo water is higher than tha t of the Catahoul~ water.

The presence of the fluoride ion and of boron (expressed as the element instead of one or more ions) is related to the tuffaceous nature of the sedi- ments. Fluoride may be derived from the fluorine-bearing minerals apat i te

DIAGENETIC ALT]~RATION AND SOIL-]~0RNIING PROCESSES 31

and fluorite, but probably larger quantities are released more easily from volcanic glass. None of the water samples from the Karnes area have unusually high F, but many contain 1 to 3 ppm whereas normal surface waters commonly contain less than 1 ppm. Likewise none of the samples are exceed- ingly high in boron, but slightly more than half of the samples tested contained more than 1 ppm, and 30 percent contained more than 2 ppm, with the highest at 11 ppm. Most water from nonvolcanic terranes contains less than 1 ppm.

In summary, the composition of the ground waters in the Jackson and Catahoula sediments clearly reflects the influence of the alteration of the volcanic detritus by their high content of silica, alkalies, bicarbonate- carbonate, fluoride, and boron.

STRUCTURAL SETTING OF THE KARNES URANIUM DEPOSITS

The uranium'deposits are in the upper part of the Jackson Group in an elongate area that lies in a grabenlike structure in western I(arnes County (Fig. i). On the northwest is the Falls City fault, with the downthrown side toward the coast, and to the south and southeast are several less continuous faults, with the upthrown side toward the coast, such as the Fashing and Hobson faults (Eargle, 1959, pp. 35-38). The rocks dip generally south- eastward toward the Gulf of Mexico, except for displacements along these faults. The faults seem to be part of the Mexia fault system, and they generally are en echelon along the trend, but individual faults may extend 5-25 miles along the strike. The faults apparently were initiated late in Early Cretaceous time and movement on some has been sporadic until at least Miocene time or later (A. W. Weeks, 1945, p. 1736). Local warping of beds along the faults produced several oil structures. Oil and gas are produced from beds of Eocene age at depths of a few thousand feet and gas and distillate from limestone of Cretaceous age at depths exceeding 10,000 ft. The possibility of a relation between the deposition of uranium and the reducing environment produced by tI2S from the petroleum has been considered (Eargle and Weeks, 1961b).

The uranium deposits in the upper part of the Jackson Group are close to the unconformity with the Miocene (?) Catahoula Tuff. The intervening Frio Clay (Oligocene ?) has been eroded and the Catahoula rests on progressively older members of the Jackson toward the north. The largest deposits are in the Stones Switch Sandstone Member of the Whitsett Formation and deposits generally below ore grade are in sand beds in the Dubose Member of the Whitsett (upper par t of Jackson Group). The most favorable area in the Stones Switch is controlled by a permeability barrier in the form of a local silty clay zone down dip in the sandstone (MacKallor and Bunker, 1958) and by the less permeable Conquista Clay Member of the McElroy Formation underneath.

32 TENTH NATIOI~AL COI~FEREI~CF~ ON CLAYS AND CLAY MII~ERALS

COMPOSITION OF THE TUFFACEOUS SEDIMENTS AND OF THE URANIUM ORES

Semiquantitative spectrographic analyses have been made on ore and low-grade samples from the uranium prospect pits and on samples taken for lithologie and stratigraphie studies away from ore. These analyses are summarized in graphs (Figs. 3, 4, 5 and 6) that show on a logarithmic scale the range of abundance of each element; also the number of samples in each spectrographic bracket, the approximate limit of sensitivity of the spectro- graphic method for each element, and the average abundance of each element in sandstone, as given by Green (1959). The samples are grouped as follows: (1) formations of the Claiborne Group (Middle Eocene), (2) formations of the Jackson Group, (3) ore and low-grade samples from the Jackson Group, and (4) Catahoula Tuff. The graphs afford easy comparison of the composition of the groups of samples with one another and with average sandstone.

The Claiborne samples (Fig. 3) contain only 25 elements detected by the spectrograph, a much more restricted suite than is found in the Jackson and Catahoula samples. The Al content ranges from that of sandstone toward that of clay because the Claiborne contains some sandy clays and glauconitic sands. The Ca, Na, and K are slightly low for both sandstones and shales. The amount of glauconite is small and has no appreciable effect on the Fe and K content. The only elements notably higher than in average sandstones are Sc and Y. Their contents are more typical of shales than of sandstones. The small number of elements in the Claiborne sediments suggests that they were more highly weathered and depleted resistates than the sediments in the Jackson Group.

Semiquantitative spectrographic analyses of 33 non-ore samples from various members of the Jackson Group (Fig. 4) show As, Ce, La, Nd, Co, Me, and Sn in addition to all the elements of the Claiborne Group. This composi- tion represents a transition to the ore suite, which contains an additional 6 elements: U, P, Ge, Zn, Tl, and Cd (Fig. 5). The Jackson samples contain a larger percentage of sodium, rare earths, arsenic, phosphorus and several metals than do the Claiborne samples. These elements in the Jackson are due to its large content of unweathered volcanic ash and other igneous fragments. Although the Jackson deposition was alternating lagoonal or shallow marine and nonmarine, the intervals of emergence apparently were not long enough for appreciable weathering or leaching of material before burial by more sediment.

Both ore and non-ore suites of samples have above average Al and Na, average ]~, and low Fe and Mg, which reflect the composition of many feldspar fragments, glass shards and sand-size grains of trachytic rock. The rare earths and trace metals could have been associated with these apparently somewhat alkalic volcanics. The ore samples (Fig. 5) show increased concentration of several elements that are present in the non-ore samples, such as Ba, V, and Me, and the greatest concentration in U. Although not indicated by the graph, Mn probably has also been concentrated. In selecting samples for

I D I A G E N E T I C ~ T ] ~ I ~ A T I O ~ A I ~ D S O I L - I ~ O I ~ M I I ~ T G 1 D R O O E S S E S 33

I0000" 0 I

I

~0000 0

ZO000"O '000"0

s 0

ZO0"O 10"0

x o

f i i

I I

g,

- ~ "~ ~ .~- ~.o

o

~o ~

0'1

0"/_ o o

l o.or ~

o.o

0 ~

O'OL ! 0"001 !

ooo~ ~ ,-~

,oz ~ o

m ~ o bl~ "

Q ~

o~.

o r~ 4.~

~ ' p

34 T E ~ T H ~ATIONAL CONFEI~ENCE O~T CLAYS AND CLAY ~/~INEI~ALS

s

,LO000-O

I 000 ' 0

s

~000 "0

IO0 '0

s

ZO0"O

I0 "0

s

Zo ' o

I ' 0

s

Z'O

"1

-s

z

'01

"O3

"0s

'0~ %'C~

I0000"0

g,

~ e g ~ .; . ~

o.., ~S !

!_I

) 'Z

- - 3 ' 0 1

0"0s

YOZ

• o~ - - - _ _ ~ - - ~ '001

J ,

SO02

J OCOZ

, 300 I

- E

_~ ~ ~ . ~ ~ . �9 ~ ~ 6 8 6 ~ ~ - m ~ . ~ _ ~ • Z 2 ~ ~ . ~ h ~ ~ _ _ _ - -

o ~ a - a

O

4g~

DIAGENETIC ALTERATION AND S O I L - F O R I ~ I I N G P R O C E S S E S 36

I0000 0

s

AO00OO

Iooo'o

s

ZO00'O

I00'0

ZOO 0

ZO0:O

10"0

~0 'o

ZO'O

I'0

s

~ -~ I I_

- 1

m

hm ~ - - i

�9 , ._ ~ ~ , ~ ,~

,~ ~:~

"ol

'og ~

~ - ~ ~ o goo oo . . . . . . . - ~ _ oo ~ N ~ -

ra ~-

o

o

, .~ o

~'~

0 ~ , ~ 0

~o .~ ~ .

36 TENTH NATIONAL CONFERENCE ON (~LAYS AND CLAY IV~INERALS

spectrographic analysis, one tends to avoid a sample with conspicuous manganese oxide coatings, and this bias results in an " a p p a r e n t " lower average Mn content in the ore. Uranyl phosphates or arsenophosphates are common ore minerals, but the spectrographic limit of detection of As and P is too high for the determination of these elements except in rich samples. The volcanic glass of the Jackson could easily have supplied the phosphorus and arsenic that have combined with the uranium in the oxidized deposits. Wilcox (1959, p. 458) gives the bulk composition of three ash samples from the 1912 eruption of Mr. Katmai , Alaska, and reports P20~ as 0.22, 0.14 and 0.05 percent. He points out tha t an ash mantle several inches thick contains large absolute amounts per acre of such mineral nutrients (for plants) as potash, lime, and phosphorus. In the ore, uranium is concentrated from 0.5 ppm in average sandstone or perhaps 10 to 20 ppm in alkalic volcanic ash to about 0.25 percent or 2500 ppm ("average" ore depending in par t on mining and milling economics), a concentration factor between a few hundred and a few thousand. The airborne radioactivity survey (Moxham and Eargle, 1961) indicates tha t the Jackson and Catahoula tuffaceous rocks probably had an average content of at least 10 to 20 ppm uranium. Molybdenum has been concentrated about 100 times and vanadium about 10 times more than in average sandstone, but neither the molybdenum nor vanadium has any commercial value in the Karnes area.

The spectrographic analyses of the Catahoula samples (Fig. 6) show a close resemblance to those of the Jackson samples, and more to the Jackson non-ore because all but two of the Catahoula group of samples are non-ore samples. No caliehified Catahoula samples were included in the graph, although one sample of the uppermost part of the Catahoula close under the Oakville Sandstone contact at the t toffman prospect is quite calcareous. The Catahoula samples show 33 elements including rare earths and niobium, and the ore samples contain uranium with higher than average V, Mo, and As; the same suite as in the Jackson samples. The only elements found in Jackson samples but not found in the Catahoula samples are P, Zn, Cd, Sn, and Ge. The first four of these have a high limit of detection in the spectro- graph and were detected in only a few Jackson samples. The fifth, Ge, is commonly associated with lignitic beds which are lacking in the Catahoula. Thus, the spectrographic analyses show no significant compositional difference between the Jackson and Catahoula sediments.

O R I G I N OF T H E K A R N E S U R A N I U M D E P O S I T S

From the t ime the writers first made a reconnaissance study of the Karnes area uranium prospects in 1955, they were impressed by the environment of these shallow deposits in par t ly altered tuffaceous rocks, with alkaline carbonate ground water, lenticular character of moderately permeable and poorly permeable beds, and subhumid to semiarid climate with evidence of intense ealiehe development in the recent past.

DIAGENETIC ~LTERATIOI~ A~D ~OIL-]~OR, MING I~OCESSES 3 7

I0000"0

s

ZO000"O

I000"0

cooo'o

LO00"O

ioo'o

s

zoo'o

fo'o - - * - - -

s

~oo I] ! I

z o _ ,

z

'ol

%'o~

,~ s a.

-11- I~

I

g N

q ~ i ~ ~ �9 ~ {-~

il �9 , ~ ~_~0o

eJ

~: ro

s

~ ~o o l

O Z

- - oo i

~ oo~

o o s

oOOE

OoL

- - OOOl

, I , -6

r

o |

v.=~

be

~ a~ o.

~ P q

o "~

N

38 TENTH NATIONAL CONFERENCE ON CLAYS AND CLAY MINERALS

The possibility of tuffaceous rocks as source beds for uranium had been considered in the Dakotas by N. M. Denson, G. O. Baehman, and tI . D. Zeller (written communication, 1950) and Denson and Gill (1955), in Wyoming by Love (1952), and in the Colorado Plateau by Waters and Granger (1953). By 1955, much previously classified work on solubility of uranium as uranyl carbonate complex ions had been released and received wide attention (Bullwinkel, 1954). This was related to the carbonate leaching process in the milling of uranium ores. However, the explanation given by Garrels (1957) first emphasized how alkaline carbonate pore water develops in tuffaceous rocks. A moderately reducing and moderately alkaline solution was shown to be the most probable type of medium to transport the uranium and associated vanadium, molybdenum, and copper (Garrels, Hostetler and others, 1957). Three diagrams given by Garrels (1960, pp. 186-189) illustrate the stabili ty relations among the uranyl carbonate and oxide hydrate compounds and vanadium compounds. He pointed out (1957) tha t the

original glassy mater ia t of the tufts reacts wi th wate r as i t . . . alters, releasing bases to solution and immobilizing hydrogen ions, wi th a concomitant rise in pH. As the p H rises, the part ial pressure of COs in the intersti t ial waters drops and the tuff becomes a CO 2 sink tha t can rob the a tmosphere , or adjacent sediments, as the ease m a y be, of CO ~. I f cations derived f rom external sources a r e no t added to the system, the net result is t ha t the waters in tufts become modera te ly to s trongly alkaline and high in carbonate. I f the tuff is wate r sa tura ted , oxygen disappears, and condit ions become mildly reducing.

That carbonate-bearing solutions are excellent solvents for uranium is demonstrated in Fig. 6.32a and b in Garrels (1960) where the stability rela- tions are shown for the uranyl diearbonate UO2(COs)2(H20)~ -2 and triearbo- nate UO2(CO3)a -4 ionic species.

The Karnes area has a large quanti ty of tuffaceous material, par t ly rhyolitic but largely traehytic to andesitic, in both Jackson and Catahoula sediments. The Jackson was deposited in a terrestrial coastal plain and shaltow lagoonal environment, and the Catahoula is entirely terrestrial in the Karnes area. The present ground waters contain sodium bicarbonate- carbonate with considerable silica and many other dements derived from the volcanic material. In the past the climate was drier and caused deposition of extensive caliehe in the soil, and the ground water was probably more alkaline than it now is. The pore water and its dissolved salts have had t ime to react with the rock instead of being flushed out of the area because the evaporation is high, the rocks are not very permeable, and the recharge is small (Anders, 1960). The precipitation of uranium from the solutions is brought about either (1) by reduction to very fine-grained, low-temperature uraninite in the reducing environment of plant fragments and by H2S, or (2) by loss of CO 2 and reaction with vanadate, phosphate, arsenate, and molybdate ions in the oxidized zone to form relatively insoluble carnotite or somewhat less stable minerals of the autunite group. The soil moisture varies greatly because the rainfall is very irregular, the area being subject to occasional downpours alternating at times with prolonged and severe drought

DIAGENETIC ALTERATION AND SOIL-FORMING PROCESSES 39

and with high temperature and high evaporation rate. At intervals, an in- crease in COs would take some of the minerals back into solution.

Studies of radioactive disequilibrium (Rosholt, 1959, and written com- munication, 1960) indicate tha t intermittent migration of uranium has continued over a long period of time. Since the Karnes deposits were first known, geologists and prospectors have been aware of the fact that ore samples are commonly out of radioactive equilibrium; that is, the percentage of "equivalent u ran ium" measured radiometricatly differs from the per- centage of uranium determined by chemical analysis. The disequilibrium is caused by leaching and migration of uranium or one or more of the daughter products. Results of radiochemical analyses of several key daughter products in two vertical suites of samples from the Korzckwa prospect trench and one from the Gembler open pit mine (J. N. Rosholt, writ ten communication, 1960) showed that some uranium deposited more than 240,000 years ago has been leached and transported along joints and bedding plane fractures into adjacent beds where it g ives" apparent ages" ranging down t o a few thousand years. One sample in pyritic dark gray silty clay just beneath the oxidized ore sand in the Gembler pit contained minute aggregates of very fine-grained uraninite associated with plant fragments. The hand sample as a whole indicated tha t m'anium had been deposited within the last 14,000 years and the uranium in carbonaceous material separated from the sample gave an apparent age of 3000 years. In other words, the only uraninite found in the Karnes deposits (by August 1961) gives the youngest age determined, and indicates downward migration from the oxidized "ore-sand" bed to the unoxidized pyritic silty clay underneath, about 40 ft below the surface of the ground.

A survey of the "equivalent u ran ium" and chemical uranium determina- tions on more than 530 samples ranging from high-grade ore down to 0.001 percent uranium indicates tha t the equivalent uranium is about 5 percent higher than the chemical uranium. This deficiency of uranium in comparison to daughter products suggests some loss of uranium from the area that was sampled, an area tha t is wide in lateral extent but not more than 40 ft deep. The " lo s t " uranium has either gone down dip in more permeable beds or been carried off in surface water to streams. The first alternative seems the more likely.

Deposits in Dura l County about 80 miles to the southwest are closely related to the Karnes deposits. A much more continuous caliche cover remains in Duval County and several uranium prospects have been found in par t ly altered tuffaceous rock just under the ealiehe. Locally traces of uranyl vanadates were found directly at the surface in exposed silicified coarse sandstone. Of interest is the largest deposit known in Dura l County where uranium occurs at a depth of about 325 ft in a reducing environment above the caproek of Palangana salt dome (Weeks and Eargle, 1960). I t is believed tha t alkaline carbonate water carrying uranyl carbonate complex ions in solution migrated down dip and deposited the uranium by reduction to uraninite in the reducing environment of the tt2S that rises from the sulfurous caprock.

40 TENTH NATIONAL CONI,~ERENCE ON CLAYS AND CLAY MINERALS

Thus, the origin of the u ran ium deposits in the Texas Coastal P la in seems to have been controlled by the complex diagenesis of these highly react ive tuffaceous source rocks, by the development of a " b u i l t - i n " solvent in the alkaline carbonate pore water, by favorable climate, and by geologic s t ructural conditions and permeabi l i ty barriers tha t allowed concentrat ion and deposi- t ion of the u ran ium, rather t h a n di lu t ion and dispersal of it .

A C K N O W L E D G M E N T S

Thanks are extended to several of our colleagues on the U.S. Geological Survey: R. M. Moxham, who invest igated the airborne radioact iv i ty of the Karnes area (Moxham and Eargle, 1961) ; J . A. MacKallor and C. M. Bunker , who studied the ore deposits of the Tordil la Hil l-Deweesvil le area (MacKallor and Bunker , 1958); J . N. t~osholt, Jr . , who studied the radioactive dis- equi l ibr ium of some of our samples; R. B. Anders, who studied the ground- water geology of the area (Anders, 1960) ; and C. L. Waring, Joseph Haffty, I t . W. Wor th ing and K. V. Hazel, who made the spectrographic analyses used in compiling the graphs in Figs. 3-6. Thanks are also grateful ly extended to S. 1%. Steinhauser, P. C. de Vergie and H. S, Stafford of the U.S. Atomic Energy Commission office in Aust in , Texas, with whom we discussed various phases of our work; to personnel of the San Antonio Mining Company and the Susquehanna-Western , Inc. ; and to m a n y local residents of the Karnes areD~.

REFERENCES

Anders, R. B. (1957) Ground-water geology of Wilson County, Texas: Texas Board of Water Engineers, Bull. 5710, 62 pp.

Anders, R. B. (1960) Ground-water geology of Karnes County, Texas: Texas Board of Water Engineers, Bull. 6007, 107 pp.

Anders, R. B. and Baker, E. T., Jr. (1961) Ground-water geology of Live Oak County, Texas: Texas Board of Water Engineers, Bull. 6105, 119 pp.

Bullwinkel, E. P. (1954) The chemistry of uranium in carbonate solutions: U.S. Atomic Energy Comm., RMO-2614, 66 pp.

Dapples, E. C. (1959) The behavior of silica in diagenesis: in Silica in Sediments, a Symposium with Discussions, See. Econ. Paleontologists and Mineralogists Spec. Pub. 7, pp. 36-54.

Denson, N. M. and Gill, J. R. (1955) Uranium-bearing lignite and its relation to volcanic tufts in eastern Montana and North and South Dakota: in Contributions to the Geology of Uranium and Thorium for the United Nations International Conference on Peaceful Uses of Atomic Energy, Geneva, Switzerland, 1955, pp. 413-418. U.S. Geol. Survey Prof. Paper 300.

Eargle, D. I/. (1959) Sedimentation and structure, Jackson group, south-central Texas: Gulf Coast Assoc. Geol. Socs. Trans., v. 9, pp. 31-39.

Eargle, D. ]:L and Weeks, A. D. (1961a) Uranium-bearing clays and tufts of south-central Texas, in Field Excursion Central Texas, 10th National Clay Conference, Oct. 14-15, 1961, Texas Bur. Econ. Geol., Guidebook no. 3, pp. 19-44.

Eargle, D. ]:[. and Weeks, A. D. (1961b) Possible relation between hydrogen sulfide-bear- ing hydrocarbons in fault-line oil fields and uranium deposits in the southeast Texas Coastal Plain : in Short Papers in the Geological Ncie~ees : U.S. Geol. Survey Prof. Paper 424-D, art. 295, pp. D-7-9.

~)IAGENETIC ALTERATION AND ~OIL-FORMING PROCESSES 41

Garrels, ~ . M. (1957) Geochemistry of sandstone-type uranium deposits: in Dunning, J. 1%. and Prentice, B. 1~., eds., Advances in Nuclear Engineering, v. 2, Nuclear Engin- eering and Sci. Cong., 2nd, Philadelphia 1957, Prec., Pergamon Press, New York, pp. 288-293.

Garrels, 1~. M. (1960) Mineral Equilibria : Harper, New York, 254 pp. Garrels, 1~. M., Hostetler, P. B., Christ, C. L. and Weeks, A. D. (1957) Stability of

uranium, vanadium, copper, and molybdenum minerals in natural waters at low temperatures and pressures (abs.): Geol. Soc. America Bull., v. 68, p. 1732.

Green, Jack (1959) Geochemical table of the elements for 1959: Geol. Soc. America Bull., v. 70, pp. 1127-1184.

Hem, J. D. (1959) Study and interpretation of the chemical characteristics of natural water: U.S. Geol. Survey Water-Supply Paper 1473, 269 pp.

Love, J. D. (1952) Preliminary report on uranium deposits in the Pumpkin Buttes area, Powder River basin, Wyoming: U.S. Geol. Survey Circ. 176, 37 pp.

MaeKallor, J . A. and Bunker, C. M. (1958) Ore controls in the Karnes County uranium area, Texas (abs.): Geol. Soc. America Bull., v. 69, p. 1607.

Moxham, ]~. 1~. and Eargle, D. H. (1961) Airborne radioactivity and geologic map of the Coastal Plain area of southeast Texas: U.S. Geol. Survey Geophys. Inv. Map GP-198.

Pettijohn, F. J. (1957) Sedimentary Rocks, 2nd ed.: Harper, New York, 718 pp. Rosholt, J. N., Jr. (1959) Natural radioactive disequilibrium of the uranium series: U.S.

Geol. Survey Bull. 1084-A, pp. 1-30. Waters, A. C. and Granger, I-I. C. (1953) Volcanic debris in uraniferous sandstone and its

possible bearing on the origin and precipitation of uranium : U.S. Geol. Survey Circ. 224, 26 pp.

Weeks, A. D. and Earglc, D. H. (1960) Uranium at Palangana salt dome, Duval County, Texas: in Short Papers in the Geological Sciences, U.S. Geol. Survey Prof. Paper 400-B, art. 24, pp. B-48-B-52.

Weeks, A. D., Truesdcll, A. H. and Haffty, Joseph (1957) l~ature of the ore boundary and its relation to diagenesis and mineralization, Uravan district, Colorado (abs.): Geol. Soc. America Bull., v. 68, pp. 1810-1811.

Weeks, A. W. (1945) Balcones, Luling, and Mexia fault zones in Texas: Amer. Assoc. Petroleum Geologists Bull., v. 29, pp. 1733-1737.

Wilcox, 1~. E. (1959) Some effects of recent volcanic ash falls with especial reference to Alaska: U.S. Geol. Survey Bull. 1028-N, pp. 409-476.